Signal Management

An introduction to signal management

Signal management is a key activity in pharmacovigilance, allowing the identification of new risks associated with a medicine. Manufacturers of medicines and regulatory agencies are constantly managing signals. But what is a signal and how are they managed?

Definition:

There are lots of definitions of “safety signal” you can find on official regulatory agency websites. For example the EMA website states:

A safety signal is information on a new or known adverse event that may be caused by a medicine and requires further investigation.

All definitions have a common theme: the “signal” is the information that could suggest a new risk with a medicine. The presence of a signal does not mean there is a definite link between an adverse event (AE) and a medicine, but after investigation we may find sufficient evidence to decide this is a new risk.

Finding signals:

The first step in signal management is detection.

Qualitative methods to detect signals include reviewing the scientific literature for individual case safety reports (ICSRs) or other descriptions of AEs with medicines. This could be the first time a new risk is described, or the article could provide new information on the known safety profile. This is how many adverse drug reactions (ADRs) were identified in the past, which is why all pharmacovigilance departments continually search the literature.

Quantitative methods use statistical analysis methods to identify signals. These complex algorithms compare the rate of reporting of AEs and medicines within a dataset. One of the ways to detect signals is to compare the rates of “Drug event pairs”. A “pair” is an AE reported with a medicine in a database. The algorithm will calculate the rate of each pair in that database and give them a ranking. If any of those pairs is more than a certain rank it will qualify as a signal for further investigation. The cut-off level for that rank can be decided by the person performing the method and regulatory agencies normally use a level of around 2, which means that if the Drug-event pair is observed more than twice as often as other Drug-event pairs it qualifies as a signal.

Validation and confirmation:

Once a signal is detected it needs to be validated, to confirm it is a real signal. This involves evaluating data to verify causal association, checking if the signal is new information and whether the event is already reflected in the product information. For example, if the signal simply repeats information that is already adequately described in the summary of product characteristics (SmPC) in the EU or the Product Information (PI) in the US, it is not new information and needs no further action.

Confirmation is not a full assessment of the signal but a process to justify further analysis. If a signal is confirmed, it is then evaluated by pharmacovigilance specialists.

Signal evaluation:

This process involves reviewing all the possible sources of data for the AE. These include ICSRs from the database; information in the scientific literature; the SmPC or PI of other products in the same therapeutic class to identify any “class effects”; non-clinical data and data from clinical trials.

The information is collected in a document and discussed by senior pharmacovigilance experts.

Taking action:

If there is sufficient evidence to conclude that this is a new risk (or new information on a known risk) then the experts will decide to update the product information. In the case that the manufacturer identifies a new risk they will submit the information to the regulatory agency and request an update to the SmPC or PI. If the regulatory agency identifies a possible new risk they can take action themselves or invite the manufacturer to take action.

However, there may be sufficient evidence to refute this signal and conclude there needs to be no change to the product information at this time. This signal will not be looked at again unless there is new information.

Or there may be insufficient data on which to base a causal assessment at this time, and the experts may decide to collect additional information and review the signal again in the future. To facilitate this they may send targeted questionnaires to collect follow up details from reporters of AEs.

Documentation:

It is very important to keep track of all the signals that have been detected and reviewed. Documenting the outcomes of all of these processes is critical so that the evidence can be shown in the event of a regulatory inspection.

To find out more here is a link to the EMA signal management web page.

Read More Pharmora Blog Posts Below

Click here to go to Pharmora’s Home Page